Performance parameters of ApsaraDB RDS-ApsaraDB RDS(RDS)-阿里云帮助中心

ApsaraDB RDS exposes performance metrics through the DescribeDBInstancePerformance API. Use these parameters to query instance metrics programmatically or to interpret data in the ApsaraDB RDS console monitoring charts.

The following tables list all supported parameters by database engine. Each parameter corresponds to one or more sub-metrics returned by the API.

ApsaraDB RDS for MySQL and ApsaraDB RDS for MariaDB

Note

MySQL and MariaDB use the same parameters.

Parameter	Description	Unit
`MySQL_NetworkTraffic`	Average inbound traffic per second and average outbound traffic per second	KB/s
`MySQL_QPSTPS`	Average SQL statements executed per second (QPS) and average transactions executed per second (TPS)	—
`MySQL_Sessions`	Active connections and total connections	connections
`MySQL_InnoDBBufferRatio`	Read hit ratio, usage, and dirty ratio of the InnoDB buffer pool	%
`MySQL_InnoDBDataReadWriten`	Average data read by InnoDB per second and average data written by InnoDB per second	KB
`MySQL_InnoDBLogRequests`	Average reads from the InnoDB buffer pool per second and average writes to the InnoDB buffer pool per second	ops/s
`MySQL_InnoDBLogWrites`	Average log write requests per second, average physical writes to log files per second, and average `fsync()` calls to log files per second	ops/s
`MySQL_TempDiskTableCreates`	Temporary tables created on disk during statement execution	tables
`MySQL_MyISAMKeyBufferRatio`	Average usage, read hit ratio, and write hit ratio of the MyISAM key buffer	%
`MySQL_MyISAMKeyReadWrites`	Average MyISAM reads from the buffer pool per second, writes to the buffer pool per second, reads from disk per second, and writes to disk per second	ops/s
`MySQL_COMDML`	Average executions per second of DELETE, INSERT, INSERT_SELECT, REPLACE, REPLACE_SELECT, SELECT, and UPDATE statements	ops/s
`MySQL_RowDML`	Average rows read, updated, deleted, and inserted in InnoDB tables per second, and average physical writes to log files per second	rows/s or count/s
`MySQL_MemCpuUsage`	CPU utilization and memory usage of a non-serverless instance at the OS level	%
`MySQL_RCU_MemCpuUsage`	CPU utilization and memory usage of a serverless instance at the OS level	%
`MySQL_IOPS`	Input/output operations per second (IOPS)	ops/s
`MySQL_DetailedSpaceUsage`	Disk usage breakdown: total, data files, log files, temporary files, and system files	MB
`slavestat`	Replication lag to the read-only instance	seconds
`MySQL_ThreadStatus`	Active threads and connected threads	threads
`MySQL_ReplicationDelay`	Replication lag to the secondary instance	seconds
`MySQL_ReplicationThread`	I/O thread status and SQL thread status. `1` = running normally; `0` = thread lost	—
`MySQL_ROW_LOCK`	Maximum row lock wait time (ms), average row lock wait time (ms), and total number of row lock waits	—
`MySQL_SelectScan`	Number of full table scans	times
`MySQL_MBPS`	Read and write throughput per second	bytes/s

ApsaraDB RDS for SQL Server

For the parameters supported by ApsaraDB RDS for SQL Server, see Parameters that can be modified for ApsaraDB RDS for SQL Server.

ApsaraDB RDS for PostgreSQL

ApsaraDB RDS for PostgreSQL supports more granular metrics through Enhanced Monitoring. The table below maps each API parameter to its corresponding enhanced monitoring metric. To view Enhanced Monitoring data in the console, see View the Enhanced Monitoring metrics of an ApsaraDB RDS for PostgreSQL instance.

Parameter	Description	Unit	Enhanced monitoring metric
`MemoryUsage`	Memory usage	%	`os.mem_usage.total.avg`
`CpuUsage`	CPU utilization	%	`os.cpu_usage.total.avg`
`PgSQL_SpaceUsage`	Disk usage	bytes	`os.fs_size.wal_dir.avg` + `os.fs_size.base_dir.avg` (unit: MB)
`PgSQL_IOPS`	IOPS	ops/s	(`os.iops.data.avg` + `os.iops.wal.avg`) / `os.iops.total.avg` × 100
`PgSQL_Session`	Total connections	connections	`db.connections.total.avg`
`PolarDBConnections`	Connection breakdown — see sub-parameters below	—	See below
`PolarDBRowDML`	Row-level DML activity — see sub-parameters below	—	See below
`PolarDBQPSTPS`	Transaction and query throughput — see sub-parameters below	—	See below
`PolarDBSwellTime`	The point in time before which the generated dead tuples can be recycled by the RDS instance	seconds	`db.swell.swell_time.avg`
`PolarDBCPU`	CPU utilization breakdown	—	`os.cpu_usage.sys.avg`, `os.cpu_usage.user.avg`, `os.cpu_usage.total.avg`
`PolarDBMemory`	Memory usage	—	`os.mem_usage.total.avg`
`PolarDBReplication`	Replication lag — see sub-parameters below	—	—
`PolarDBLongSQL`	Slow SQL statements by duration threshold — see sub-parameters below	—	See below
`PolarDBLongIdleTransaction`	Idle transactions by duration threshold — see sub-parameters below	—	See below
`PolarDBLongTransaction`	Long-running active transactions — see sub-parameters below	—	See below
`PolarDBLongTwoPCTransaction`	Long-running two-phase commit transactions — see sub-parameters below	—	See below
`PolarDBLocalIOSTAT`	Data disk I/O — see sub-parameters below	—	See below
`PolarDBLocalDiskUsage`	Disk space by directory — see sub-parameters below	—	See below

PolarDBConnections sub-parameters

Sub-parameter	Description	Enhanced monitoring metric
`mean_active_session`	Active connections	`db.connections.active.avg`
`mean_idle_connection`	Idle connections	`db.connections.idle.avg`
`mean_total_session`	Total connections	`db.connections.total.avg`
`mean_waiting_connection`	Connections in the waiting state	`db.connections.waiting.avg`

PolarDBRowDML sub-parameters

Sub-parameter	Description	Enhanced monitoring metric
`mean_tup_returned_delta`	Rows returned per second by full table scans	`db.sql.tup_returned.avg`
`mean_tup_fetched_delta`	Tables returned per second by index scans	`db.sql.tup_fetched.avg`
`mean_tup_inserted_delta`	Rows inserted per second	`db.sql.tup_inserted.avg`
`mean_tup_updated_delta`	Rows updated per second	`db.sql.tup_updated.avg`
`mean_tup_deleted_delta`	Rows deleted per second	`db.sql.tup_deleted.avg`

PolarDBQPSTPS sub-parameters

Sub-parameter	Description	Enhanced monitoring metric
`mean_commits_delta`	Transactions committed per second	`db.transactions.xact_commit.avg`
`mean_rollbacks_delta`	Transactions rolled back per second	`db.transactions.xact_rollback.avg`
`mean_deadlocks_delta`	Deadlocks per second	Number of deadlocks
`mean_tps`	Total transactions per second (commits + rollbacks)	`db.transactions.xact_commit.avg` + `db.transactions.xact_rollback.avg`

PolarDBReplication sub-parameters

Sub-parameter	Description	Unit
`mean_replay_latency_in_mb`	Replication lag on the secondary instance (data not yet replayed)	bytes
`mean_send_latency_in_mb`	Replication lag from the primary to the secondary instance (data not yet sent)	bytes
`mean_logical_rep_latency_in_mb`	Logical replication lag	bytes

PolarDBLongSQL sub-parameters

Sub-parameter	Description	Enhanced monitoring metric
`mean_one_second_executing_sqls`	SQL statements running for more than 1 second	`db.slow_sql.one_second.avg`
`mean_three_seconds_executing_sqls`	SQL statements running for more than 3 seconds	`db.slow_sql.three_seconds.avg`
`mean_five_seconds_executing_sqls`	SQL statements running for more than 5 seconds	`db.slow_sql.five_seconds.avg`

PolarDBLongIdleTransaction sub-parameters

Sub-parameter	Description	Enhanced monitoring metric
`mean_one_second_idle_transactions`	Transactions idle for more than 1 second	`db.long_transactions.idle_one_second.avg`
`mean_three_seconds_idle_transactions`	Transactions idle for more than 3 seconds	`db.long_transactions.idle_three_seconds`
`mean_five_seconds_idle_transactions`	Transactions idle for more than 5 seconds	`db.long_transactions.idle_five_seconds.avg`

PolarDBLongTransaction sub-parameters

Sub-parameter	Description	Enhanced monitoring metric
`mean_one_second_transactions`	Active transactions running for more than 1 second	`db.long_transactions.active_one_second.avg`
`mean_three_seconds_transactions`	Active transactions running for more than 3 seconds	`db.long_transactions.active_three_seconds.avg`

PolarDBLongTwoPCTransaction sub-parameters

Sub-parameter	Description	Enhanced monitoring metric
`mean_one_second_two_pc_transactions`	Two-phase commit transactions running for more than 1 second	`db.long_transactions.two_pc_one_second.avg`
`mean_three_seconds_two_pc_transactions`	Two-phase commit transactions running for more than 3 seconds	`db.long_transactions.two_pc_three_seconds.avg`
`mean_five_seconds_two_pc_transactions`	Two-phase commit transactions running for more than 5 seconds	`db.long_transactions.two_pc_five_seconds.avg`

PolarDBLocalIOSTAT sub-parameters

Sub-parameter	Description	Enhanced monitoring metric
`mean_local_iops_read`	Read operations from the data disk per second	`os.iops.read.avg`
`mean_local_iops_write`	Write operations to the data disk per second	`os.iops.write.avg`

PolarDBLocalDiskUsage sub-parameters

Sub-parameter	Description	Unit	Enhanced monitoring metric
`mean_local_pg_wal_dir_size`	Disk space used by write-ahead log (WAL) files	MB	`os.fs_size.wal_dir.avg`
`mean_local_pg_log_dir_size`	Disk space used by log files	MB	`os.fs_size.log_dir.avg`
`mean_local_base_dir_size`	Disk space used by data files	MB	`os.fs_size.base_dir.avg`